Effect of statins in preventing hospitalizations for infections: A population study

• Published in: Pharmacoepidemiology and drug safety 2018-08, Vol.27 (8), p.878-884
• Main Authors: Policardo, Laura, Seghieri, Giuseppe, Gualdani, Elisa, Franconi, Flavia
• Format: Article
• Language: English
• Subjects: Bacterial infections; Diabetes mellitus; Health risks; hospitalizations; Infections; Infectious diseases; Ischemia; number needed to treat; pharmacoepidemiology; Population; Population studies; population study; propensity score; Statins
• ISSN: 1053-8569; 1099-1557
• DOI: 10.1002/pds.4557

Purpose To investigate whether the hospitalization rate for bacterial infections was modified by statin therapy in a population retrospectively followed up, over years 2011 to 2015. Methods By using administrative databases, the 5‐year hospitalization rate due to bacterial infections in population living in Tuscany, Italy exposed to statin therapy (n = 52,049) was stratified by 5 prescribed daily doses classes (0%‐20%, 20%‐50%, 50%‐80%, 80%‐100%, ≥100% of DDD) and subsequently compared with that of a population of untreated individuals (n = 3 300 ,675), matched through a propensity score accounting for all available covariates potentially able to modulate risk of infections such as age, gender, previous hospitalizations for infections, cardiovascular events, previous co‐morbidities, diabetes, as well as general practitioners' proactive behaviour of care delivery according to current guidelines. Results Unmatched individuals of each treatment‐class had significantly more hospitalizations than controls, while matched treated people, apart from those in class 0% to 20%, had a decrease of hospitalizations, as large as the increase in prescribed drug. Statin effect in reducing hospitalizations translated into a number needed to treat (NNT) ranging across treatment strata from 102 to 54. Conclusions Compliance to statin prescribed daily doses above the threshold 20% of DDD, along a 5‐year follow‐up, prevented hospitalizations due to infectious diseases in a large unselected population, after adjusting for covariates able to modulate baseline risk of infections. The NNTs to avoid 1 hospitalization for infections resulted on average not too dissimilar from a value lying between the 95% CI of NNTs previously found for primary prevention of 1 incident coronary ischemic event (72 to 119).